Cover photo credit: Albertus Eskes.
This is a summary of the research by Eskes et al., published in the journal Agrotropica in 2018. All tables and figures in this summary belong to the researchers involved.
What is cacao pulp?
It is the fruit of the tree, Theobroma cacao. To be more specific, we are talking about the fleshy portion of the fruit, which we call the pulp. In this summary, the term “fruit” will refer to the fleshy fruit portion and will be used interchangeably with the term “pulp.” This fleshy fruit is held within the pod husk, and surrounds each individual seed. The cacao pulp is mucilaginous and composed of spongy parenchyma cells. The pulp is mostly water, as well as 10-13% sugar, 8-10% salts, 2-3% pentosans (polymers of simple sugars), and 1-2% citric acid.
Cacao pulp is mainly used as a substrate for fermentation of the seeds. It’s difficult to remove the seeds from the pulp, but during fermentation the decomposing fruit essentially “melts away” from the seed, so that they can be dried, roasted, and turned into chocolate. The fermentation process also breaks down proteins and carbohydrates within the cotyledon (or kernel) of the seed, which is an important step for building the chocolate flavour later on during the roasting stage.
The flavour of the cacao pulp can vary between trees of different genetic varieties, and can even differ between individual trees with similar genotypes. The authors suggest that although the majority of genotypes appear to have lower levels of pulp flavour and intensities (most bulk cacao trees produce pulp that isn’t very flavourful), there are some genotypes that have long been known to have extremely flavourful and aromatic pulp (such as strains SCA6 and EET62). The reason so much cacao today lacks much flavour is partly due to selecting trees for yield and resistance to disease, while neglecting flavour.
This is what has lead to the creation of the infamous CCN-51 strain, developed in Ecuador in the 1980s. Much more profitable for farmers to grow, but very low quality when it comes to pulp and cacao flavour. The authors have suggested that moving forward, growers and researchers should select for yield and disease resistance from genotypes that have already been found to be flavourful.
Relationship between pulp flavour and fine flavour chocolate:
Many people today place a great importance on terroir and soil as the origin of the fine complex flavours in chocolate. However, there is a great deal of more evidence which supports the idea that flavour within cacao stems from the genetics of the tree and the flavour of its pulp, with little to no evidence supporting the idea that soil impacts flavour.
Research from this study has unveiled 64 unique cocoa pulp and bean sensory profiles (combinations of flavours and aromas) in the Chuncho variety found in Peru. This is likely due to the generations of farmers who selected for trees that produced flavourful juices. This strengthens the idea that certain strains of cacao contain much more genetic variety in regards to flavour than is currently accepted.
In 2006, the first author (Albertus Eskes) hypothesized this relationship based on his experience tasting the pulp of various types of cacao trees in 11 different cacao producing countries.
In 2010, Pino et al. were the first to identify volatile organic compounds (VOCs) in fresh cacao pulp from a farm in Colombia. VOCs can also be understood as the aroma molecules, giving foods their distinct flavours beyond that of the 5 tastants (bitter, salty, sour, sweet, and umami).
Kadow et al. (2013) compared VOCs of the pulp of 3 genotypes, one being bulk and 2 being fine. Coincidently the flavour (or VOC concentration) for the pulp from the bulk CCN-51 cacao was very low, while the flavour (VOC concentration) for the other two fine cacao types were very high. So there the idea that there is a relationship between pulp flavour and cacao flavour is beginning to emerge.
Based on Kadow’s work, Hegmann (2015) analyzed the VOCs of 5 genotypes selected at CATIE, in Costa Rica. The quantities of these VOCs differed according to genotype, season, and fruit ripeness. This means that choosing unripe fruit, or season does impact fruit flavour which impacts seed flavour, which impacts chocolate aromas.
Barry (2009) states that these aromas are species-specific, and key for attracting frugivores (including humans).
The authors in this paper state that their work here is the first systematic study in T. cacao to demonstrate the feasibility of selecting for flavour in single-tree genotypes by using the taste of the pulp and raw bean as the selecting factor. This practise is selecting for the fruit likely came from millions of years of frugivore selection.
Frugivore and human pulp preference vs Chuncho
Cacao depends on frugivores to take the fruit. The cacao pod does not fall and open on its own (indehiscent), but needs to be picked off. This has lead to a strong evolutionary relationship between cacao and frugivores. Frugivores were choosing cacao based on their very specific preferences, which likely lead to the wide variety of flavours identified in Chuncho cacao.
In the La Convención region of Peru, there are 5 frugivore animal families or genuses observed by researchers (the spider monkey, squirrel monkey, coatis, rodents, and squirrels), as well as the Matsigenkas people, who were associated with cacao even before the Inca Empire.
Colour, smell, and texture are the most important traits to attract frugivores. It’s suggested once a frugivore identifies the pulp of a fruit as sweet and nutritious, that’s enough to ensure it gets eaten. It’s suggested that aroma attract the frugivores first, and then flavourful pulp stimulate eating and seed dispersal. It’s also possible humans, including the Matsigenkas and today’s Chuncho growers, selected trees for their pulp more than for their odours.
Chuncho pulp and beans mimic flavours of 4 fruit species which grow near T. cacao, including soursop, custard apple, other annona species, and inga species. Frugivores also consume these fruits. This could have lead these frugivores to select for cacao trees whos pulp mimic the same flavours found in other fruit species they already learned to consume.
The Chuncho cacao variety of Peru
“Chuncho” means “from the jungle”. This cacao is native to La Convencion province (Urubamba valley) in the Cusco region of Peru. Reports exist of a cacao bean trade by native Matsigenka people of Peru with highlanders before and during the Inca Empire (Johnson and Johnson, 1996).
From a genetic perspective, Chuncho is part of the Contamana group of cacao varieties, one of 10 wild varieties suggested by J. C. Motamayor. This includes the the cultivar SCA6. SNP marker studies show Chuncho to have a closer genetic relationship with that of cacao from Madre de Dios and Beni river basin in Peru. That said, Chuncho appears to have a very strong genetic identity on its own. The molecular diversity within the Chuncho variety can be considered moderate to high compared to other native cocoa varieties (Zhang, 2014).
Chuncho cacao has been cultivated since at least the 16th Century. The Matsigenkas not only grew the cacao, but also ate the pulp and consumed its slightly fermented & roasted beans. Paz Soldan (1852) stated Chuncho cacao from the Echerate Estate as “the best cocoa in the world”. Cultivation of Chuncho increased in the 19th Century, and by the 1980s they were producing about 10,000 tons. However, more recent production is closer to 2600 tons due to limiting factors such as low productivity, aging trees, and the competition of bulk cacao replacement crops. Today, Chuncho is mainly used as a low-valued bulk cacao for cocoa butter and cocoa powder production. However, we will see from results in this study that the Chuncho variety can clearly be included as a fine cacao variety and developed into a commercially available fine cacao variety
What’s interesting is that today, Chuncho farmers eat the pulp from select trees around them. They identify individual trees that produce very unique delicious fruit. Different individual trees will have different flavour profiles. This cacao fruit eating of very specific trees is not a behaviour known or common elsewhere in the world. Growers from other parts of the world don’t make a habit of selecting for and consuming the cacao fruit of individual trees. This study focused on the idea that perhaps these specific trees with their unique and delicious flavours may also produce cacao and chocolate with unique and delicious flavours.
Cacao beans today need to be fermented and roasted to rid them of the high levels of acidity, bitterness, and astringency found in most cacao, and especially in bulk cacao. However, the authors observed that certain Chuncho beans, even when unfermented or fermented very little, are neither acidic, bitter, nor astringent. They did observe that fermenting this cacao for more than 4 days would increase the levels of those non favorable attributes (acidity, bitterness, and astringency), meaning Chuncho seems to need little to no processing to produce flavourful cacao beans.
Objectives For Paper:
Six sensory traits were observed, which included five pulp traits (acidity, bitterness, astringency, sweetness, and flavours/aromas), as well as one bean trait which included bitterness.
The authors wanted to measure the level of intensity for the above 6 traits through tasting and smelling the pulp and cacao beans. Although many studies before this included measuring the levels of VOCs in pulp and beans, here the authors wanted to conduct their findings in a more practical way. If this method was successful, it could easily be taught to cacao farmers who could assess the flavours and quality of trees themselves without the use of expensive lab equipment.
The researchers focused mainly on the Chuncho cacao trees of Peru, but also analyzed and compared results with fruit pulp from trees outside this geographic region as well.
1. Nine Genotypes & their pulp sensory trait associations in Tingo Maria, Peru, 2007.
Table one displays sensory traits for 4 fine-flavoured cacao. You can see that the flavours within the fruit are not identical to those that appear in the cacao, although they are related. For instance, the grape flavour in the SCA6 clone pulp becomes brown fruit or raisin flavour when that same seed is turned into chocolate. The jasmine aroma in the Nacional pulp becomes floral in the chocolate made from those seeds. This observation will be useful later on when analyzing the Chuncho cacao varieties.
The pulps of 9 genotypes evaluated in Tingo Maria, Peru in 2008, were analyzed and judged according to what panelists preferred. Pulps which were most preferred were also the pulps that contained higher intensities of odour, flavour, and sweetness. The least preferred pulps were also associated to acidity and astringency, although not statistically significant.
Cacao containing pulp which was rated higher for acidity and astringency also contained beans that were more bitter and astringent, all of which were negatively correlated with pulp sweetness.
The most preferred pulp of these 9 came from the Ecuadorian clone EET 62, which contained sweet pulp, and strong odour and flavour. The second most preferred was the variety coined Pandora 1, which contained a strong soursop flavour. Soursop is a fruit native to the Americas, but also grown in Southeast Asia. The third preferred was ICS 1, a trinitario clone, which contained strong fresh fruit and lemon flavour. The least preferred was CCN 51, a bulk cacao also from Ecuador which was developed to be resistant to disease, but is also known for its poor quality. The panelists found the pulp of CCN 51 to be acidic and astringent, with bitter beans, and no flavour or aroma. CCN 51 appeared to be the extreme opposite of EET 62.
2. Identification of 12 unknown cocoa varietal sensory traits in 7 different locations outside La Convencion, 2007-2017
Cacao pulp assessments were carried out in 7 locations within Trinidad, Peru, Ecuador, and Brazil over a 10 year time frame by various scientists. The researchers assessed the main flavour and aroma traits within the cacao pulp samples encountered in the regions they visited. The table below includes the flavours and the type of cacao associated with it.
According to data from Table 3, the most common pulp flavour was soursop, found in all 7 locations.
Five flavours including banana, soursop, jasmine, citrus, and annona (aka sweetsop) were found in Amazon genotypes from Peru.
In Brazil, the mango/rose flavour profile was found in the EET397 clone that originated in Ecuador. Also in Brazil, jasmine aroma was found in an amelonado variety. The fine cacao, Nacional from Ecuador, is well known for its jasmine aroma. It’s interesting to find this aroma within an amelonado variety since it is classified as a “bulk” cacao, and not often known for fine flavours. This may offer hope that other genotypes not known as “fine” cacao may hold some promising flavours if only they are analyzed.
The banana flavour came from trees in Trinidad. The beans from this cacao were turned into chocolate which also contained banana flavours described by cocoa specialists Majaraj, Ed Seguine, and Albertus Eskes. This continues to add evidence to idea that pulp flavours do influence the flavour of the chocolate.
Many of the flavours found in these studies were also found in Chuncho cacao in Peru.
3. A survey on pulp flavour and aroma among 226 chuncho trees in La Convencion, 2012-2017.
The following will analyze the Chuncho variety sensory traits (flavours) for pulp, raw beans, and pulp & beans combined.
Chuncho Pulp Sensory Traits:
The pulp flavours and aromas of 100 Chuncho trees were evaluated in 2012, and 97 of the 100 trees were described to have pulp flavours and aroma. These flavours and aromas were classified into 30 sensory profiles. A sensory profile is basically what the fruit/bean tasted and/or smelled like.
These sensory profiles include 17 fruit flavours and 9 aromas/spices. The most frequently encountered profiles were soursop/jasmine, floral, or floral, and citrus/jasmine identified in 13 trees.
Chuncho Raw Bean Sensory Traits:
In April of 2015, the sensory traits of 126 slightly fermented raw bean samples were analyzed.
Of the 126 samples, 92 received high preference scores of 3 and above (on a scale of 0 to 5). The researchers identified 39 flavour profiles in total for the raw beans. The most frequently encountered profile was mandarin/jasmine, soursop/floral, and cranberry/malt.
Chuncho Pulp & Raw Bean Sensory Traits Combined:
Combining these pulp and bean evaluations resulted in 64 unique flavour profiles among 226 Chuncho trees (100 for pulp, 126 for the bean). Of the 64 flavour profiles, 25 (39%) were made up of only one flavour or aroma trait. The remaining 39 (60%) are based on combinations of two or more traits. 9% are made up of 3 flavour and/or aroma combinations.
There were 18 unique pulp flavours identified (although table 4 mistakenly states 17), and 15 unique bean flavours (5 of which were the same as the pulp flavours), which gives a total of 28 unique flavours between both pulp and beans.
There were 9 unique pulp aromas, and 8 unique bean aromas (5 of which were the same as the pulp), which gave a total of 12 unique aromas between both pulp and beans.
Therefore, the total of all sensory traits including flavours and aromas for the Chuncho trees was 40 (12 + 28). Out of these 40 traits, 29 mimicked those of known fruit, flower, and spice species. It appears that these Chuncho cacao mimicked not only different fruit species, but different varieties of fruit species as well, such as the 4 different banana varieties listed in Tables 4 and 5.
4.0 Relationship between flavour and aroma identities with quantitative sensory traits in a group of 64 preferred chuncho bean samples.
20 of the 64 most preferred bean samples were further analyzed by evaluating the bitterness, acidity, astringency, fruitness, and floral, and overall preference on a scale from 0-5. It’s interesting to point out that although these Chuncho beans were not fermented very long, 80% of astringency and bitterness scores were 2 or less, and 89% of acidity scores were 2 or less, pointing out that these flavourful beans and pulps also produced seeds that were low in the less favorable sensory traits. Normally, cacao is fermented longer to help improve the flavour and reduce bitterness.
Bean flavour vs. basic flavour associations:
Statistical analysis of 64 of the 126 most preferred bean samples were conducted, comparing the bean flavour with basic flavour (bitter, acidity, astringency). 13 bean samples with mandarin/jasmine pulp flavour produced significantly more bitter beans. 8 bean samples with citrus/jasmine flavour showed higher than average acidity scores than that of the control. And those with red fruit flavours had a significantly higher astringency score.
Chuncho pulp vs bean vs chocolate:
This table identified flavours of the pulp, seeds from that pulp, and chocolate made from those seeds in Chuncho genotypes. It’s interesting to see how the progression of flavours found in fresh fruit appear to me altered in the raw beans and chocolate, yet consistent with the pulp flavour. For instance, the grape flavour in the pulp transformed into raisin in the bean and chocolate. Green apple detected in the pulp became dried apple. Jasmine and rose in the pulp became general floral notes in the chocolate. Banana in the pulp became dry banana in the bean, and banana jam in the chocolate.
Chuncho sensory trait diversity vs other fleshy fruit diversity:
We have a great appreciation for the range of apple varieties, or the numerous aroma profiles of fine wine. However, the authors break down how impressive Chuncho cacao and cacao in general stands up to these foods that we consider to have a great diversity of flavour and aroma.
Authors suggest Chuncho mimics 21 flavours and 7 aromas from other fruit species alone. In total, Chuncho mimics flavours and aromas of 29 known fruit and flower/spice species. This diversity in flavour traits seems to be quite unique to cacao. Let’s compare.
For example, Apples. There are 52 descriptors to analyze 16 apple varieties, perhaps the greatest sensory diversity among commercial fruits. However, the descriptors only refer to two other species of botanicals: pear and almond.
A descriptor list for mango varieties only contained non-specific flavours such as sweet, acidic, juiness, commonly found in many fruits, and not very unique. They didn’t refer to flavours of other species of fruits as do cacao.
Now, there is great diversity in aroma in regards to wine. However, these aromas are generated by the processes of wine making, which includes grape juice fermentation. However, the same diversity in flavour doesn’t occur at the grape level. If we are to compare fairly, we need to compare the fruit of the cacao to the grape, before any processes are applied. In this case, the variety of grape aromas are limited to muscat and shiraz. Grape juices of different grape varieties show little sensory variation, and much less than that of the pulp of cacao.
The authors suggest Chuncho could be the “mother” of all T. cacao flavour and aroma traits found thus far.
Genetic determination of Chuncho sensory profiles & future applications:
Many Chuncho sensory profiles are multi-trait profiles. For example, mandarine/jasmine sensory profile always displays strong mandarin fruit flavour and medium to strong jasmine odour combined with higher than average bean bitterness.
The soursop flavour is always associated with translucent pulp. Other multi-trait sensory profiles include: soursop/jasmine, citrus/jasmine, mint/floral, cranberry/rose, mint/jasmine, nuts/rose, and peach/nuts/rose.
The authors put the idea forward that these multi-trait associations suggest a pleiotropic dominant gene action, which means that one gene can turn on the expression of multiple unrelated traits. Therefore, T. cacao may have benefited from pleiotropy as an evolutionarily favorable mechanism associated with several traits attractive to frugivores, including humans.
Authors have shown that selecting fine flavour cocoa genotypes can be done quickly and efficiently by pulp evaluation for fine-flavour sensory traits. Farmer associations and research institutions could promote large scale competitions for farmers to identify trees within the varieties they have, and select the most delicious pulp/non bitter seed combinations.
Once the trees are identified, they could be cloned for conservation, and bred perhaps into hybrids that still showcase the favorable traits, aiming to combine not only high yield and resistance, but also fine flavour. Chuncho can be crossed with parental genotypes (e.g. PA121 and PA169) that are known to produce high yielding and resistant hybrids where natural pulp doesn’t interfere with the fine flavour of the Chuncho, which is expected to dominate the hybrid.
Fine-flavor cocoa varieties, including well known varieties such as criollo and nacional, present 6 common aroma traits (Table 1). Chuncho is unique in that it not only contains an array of favorable flavour traits, but also low on the unfavorable traits of bitterness, acidity, and astringency. Commercial cacao offers a very narrow profile, while the Chuncho variety offers 64 interesting flavour and/or aroma combinations.
The researchers have demonstrated that the Chuncho variety includes these 6 aroma traits as well as 34 other pulp sensory traits. Therefore, 34 of the Chuncho sensory traits are new in regards to fine flavour cacao. Two trees identified as Trinitario expressed banana and soursop flavours. This suggests Chuncho-like flavours can be found within trinitario and offers hope for finding them in other fine flavour cacao varieties.
As well, the research between 2007 and 2017 showed 12 different fruit flavours and aromas were detected in pulp of widely different varieties and locations (table 3). This confirms these Chuncho-like flavours are not just localized to Peru. As well, it also adds to the idea that the flavours within cacao have strong genetic factors.
Authors strongly suggest that with a system in place of single-tree pulp assessments, that it’s possible to find the numerous Chuncho flavours in other cacao varieties not yet recognized as fine. An example of this from the study is a jasmine trait identified in a Brazilian Amelonado variety in 2017 (Table 3), which is a variety that is often determined to be bulk cacao.
Finally, the researchers believe that all sensory traits identified in cacao varieties should be considered T. cacao traits rather than link specific flavour traits to to specific varieties. It has been clearly observed that many fine flavours and aromas occured in various genotypes in various geographical locations. It’s beneficial to many growers around the world for us to remove the bias associated with specific origins or genotypes in order to give their cacao a chance on the fine chocolate market. It would be a great loss to see many fine cacao genotypes go unnoticed and unmarketable because they don’t fit in with the current trends or guidelines.